The viral Epidemiology Branch has used a variety of approaches to define the nature and magnitude of HIV-1 associated malignancies, including analyses of population-based cancer registries, death certificates, prospective cohorts, and laboratory studies. A major focus in the past year has been on the nature and etiology of Kaposi's sarcoma, which was recently associated with a newly discovered herpesvirus, HHV-8. In a collaboration with the Laboratory of Pathology, NCI, HHV-8 was localized by microdissection to spindle cell areas of Kaposi's sarcoma tissue sections whereas it was absent from stromal areas. VEB researchers have extended their previous finding of monoclonality of this tumor by determining that multiple lesions share the same clonal origin, further establishing Kaposi's sarcoma as a true malignant neoplasm. In studies of HIV-infected hemophilia patients, non-Hodgkin's lymphoma incidence increased exponentially with duration of HIV infection, reaching 7% (95% confidence limits, 1%-12%) per year 15 or more years after HIV-1 seroconversion. The incidence of malignancies other than Kaposi's sarcoma and non-Hodgkin's lymphoma was not significantly increased overall, although Hodgkin's disease was increased 5.6-fold (95% confidence limits, 1.1-16). HIV-viral load was not associated with lymphoma or cancer risk. Cancer surveillance data from the pre-AIDS era were analyzed for cancer risks associated with single marital status (a surrogate for homosexuality in men); except for anal cancers, no increases were found. Recent analyses of the Multistate AIDS-Cancer Match Registry have elucidated the impact of AIDS on the epidemiology of central nervous system and systemic non-Hodgkin's lymphoma and on Kaposi's sarcoma. No increase was detectable in risk of nasopharyngeal carcinoma, an EBV-associated tumor. Current efforts have linked 90,000 AIDS cases with cancer registry data, and an expanded match is planned for 1997 to better explore rare tumor types and cancer in small population subgroups.